Programmer for MX25S6473F (MacBook A2681, U1900, SPI ROM, 1.2V power)

[baglaivlad]

Hi. Please share your knowledge, which programmer can be used to flash firmware for MX25S6473F (SPI ROM, 4x4mm 8-XSON)? This IC can be found in MacBook Air M2 (A2681), U1900. Power and data are 1V25.

https://www.mxic.com.tw/en-us/produc...Flash&n=PM2860

[mon2]

Hi. The flash device is operating @ 1v25. Would you have a suitable level shifter to make the interface pins safe to be used with a programmer? If not, it is required.

For example, a bidirectional FET based level shifter from Texas Instruments is suitable to level shift to/from the 3v3 levels found on most programmers and shift to the required 1v2 levels for this component. Respectively, this component would need to be powered by 1v2 on the conversion board. Someone out there must offer such a level shifter but if not, can share more details on how to proceed with the conversion board.

[baglaivlad]

Hi. Thank you. I got a 3.3 to 1.8 level shifter, but not for 1.2V. Unfortunately, can't find it on sale anywhere.
TXB0108PWR IC is available for purchase. Will it fit my needs? Do you have any schematic of 1.2V level shifter to built it? Thank you.

[mon2]

Can you share pics of your 1v8 level shifter? Need to check if that design can be converted to support 1v2 level shifting. Which level shifter IC is onboard this translator? Really surprised if an off-the-shelf translator is not available. Plan B is to consider the DIY board and perhaps use boards from Adafruit. They are simple to use. One side of the translator is mated to the 3v3 rail from your programmer. Other side is mated to the target voltage required by the flash device, which is 1v2 for your target memory. Then wire:

SPI_CLK
SPI_MOSI
SPI_MISO
SPI_CS

1:1

With the above configuration, the host programmer will believe it is working with a 3v3 flash device but you are level shifting to 1v2 to/from the device.

I recall a company in the UK that offered some unique boards for the macbook world but cannot remember their name at this time.

[baglaivlad]

Sure. Please see the photos attached. It is based on ALVC164245 IC and buck converter AMS1117 1.8V. I assume AMS1117 can be replaced with the one that outputs 1.2V instead of 1.8V. The question is can I replace ALVC164245 with TXB0108PWR? I should figure it out.

Thank you for your explanation! If any level shifter is available for purchase in US/China, I can buy it and wait some time for delivery. Although, I think with your help I can built my own based on TXB0108PWR.

[baglaivlad]

Actually, it seems that Nexperia ALVC164245 used in my 3.3 to 1.8 level shifter supports wide voltage range, from 1.5 to 3.6V (https://assets.nexperia.com/document...ALVC164245.pdf).

MX25S6473F (IC from MacBook) has Vcc range of 1.14V-1.6V, according to its documentation (https://www.mxic.com.tw/en-us/produc...Flash&n=PM2860).

How do you think, will it be okay to just convert my level shifter to 1.5V (by replacing ASM1117) and use it with MX25S6473F IC? Yet it should work fine with 1.5V as Vcc, I don't know it it's okay to supply 1.5V onto its logic inputs.

[mon2]

Yes, found the same on our local Amazon store. This is a bit of a pickle.

The LOWEST voltage supported by this AVLC16425 (from the datasheet) is 2v7. Technically, 1v8 is beyond spec unless the vendor they have used supports this lower voltage. Respectively, the use of 1v2 for the low side is even further away from the officially supported voltage rail for the VCC. We need to be able to use 1v2 on one side and 3v3 on the other side.

Perhaps this silicon can be 'pushed' down to use 1v2 which means just switch out the SOT223 LDO from 1v8 to a 1v2. Let me see if they have this in stock locally here in Canada for experimenting. We have 1v2 LDOs in the building. It will be a hack but the most simple test is to just use a 1v2 rail for the same piece of hardware you have.

LOL - we are almost on the same wavelength - worth testing but not sure who makes the level shifter on your board. Are you sure yours is from NXP? TI also offers these but they have a higher VCC requirement.

[baglaivlad]

🙂 Yes, TI version has higher voltage requirements. It even does not support 1.8V and meant to be used in 5V to 3V shifters. It confused me too. But I think the logo on my IC is Nexperia (or its Chinese knockoff), which supports 1.8V.

I can order both ASM1117 for 1.5V and 1.2V and try it out. With 1.5V it should work just fine, don't know about 1.2V.

According to my latest question, do you think it is safe to put 1.5V on MX25S6473F inputs? Vcc can be 1.14V-1.6V.

[mon2]

Yes, we are thinking the same here again. Yes, it is perfectly safe to feed 1v5 on your flash device if the flash device is powered from 1v5. I think it is safe and borderline ok to consider this experiment to operate the level shifter @ 1v5 on the low side and 3v3 on the high side.

I also believe you are an early adopter (bit of a pun here) on this experiment. Continue to post your results..others will want to know in the future. If really really required, we can design a low cost PCB to do the same and open source it but we are still playing catch up on some past due new designs. That pays our bills and keeps our led lights on

[baglaivlad]

Thank you! Will definitely keep you informed! Ordering ASM1117...

[mon2]

Two more comments before you proceed.

1) Double check the pinout of the SOT223 LDO that you wish to install on this adapter board. Be sure that the MIDDLE TAB is ground on the replacement LDO. They do sell in assorted pinouts. We got stuck on this during covid times and had to pay a premium for the replacements.

2) Very important, double check if the filter caps on the LDO are supporting ceramic caps or not. You have ceramic caps on your board and these are very low ESR parts. On old school but cheaper LDOs, you CANNOT use ceramic caps and this causes a ringing on the voltage rails. Nasty but true. The older LDOs will only support higher ESR parts like tantalums or electrolytics. Be sure to confirm this else the voltage will oscillate - not a good thing to have.

If in doubt, post the links of what you plan to purchase for a quick review.

PS: For one of our designs were we had to use the older style of regulators, we applied a 1 ohm to ground in series with the ceramic caps so that the total ESR was high enough so that the LDO would not create this ringing. Kooky but worked out fine.


reference reading:

How to stabilize a buck or LDO to use ceramic caps - EDN

[mon2]

In addition to the above...double check what is the voltage for Vcc being supplied by your programming tool. Is your tool modified for 3v3 on the high side Vcc? That is recommended for the Vio pins from the programmer -> then the voltage level shifter will translate to the 1v5 (after you swap out the LDO) side.

[baglaivlad]

Thank you for additional information. I will buy LD1117AS12TR (for 1.2V, https://eu.mouser.com/datasheet/2/38...7a-1849492.pdf) and AMS1117-1.5 (for 1.5V, http://www.advanced-monolithic.com/pdf/ds1117.pdf). Both of them and ASM1117 on my conversion board have the same pinout (1 - Gnd, 2 - VOUT, 3 - VIN).

As far as I understand the datasheet, LD1117AS12TR can be used with ceramic capacitors I have on my board. While about AMS1117-1.5 I am not so sure. It says about tantalum capacitors, not ceramic. Is it ok?

My programmer supplies 3.3V, I checked it.

[mon2]

1) the pinout looks good across these referenced LDOs.

2) these LDOs are not listed to be ceramic cap friendly but do see ceramic caps on your adapter board and also in the pics on our local Amazon stores for these same PCBs. If you view the output on an oscilloscope, you will most likely see a ringing of this output voltage. Tantalum caps have a different shape and also a polarity.

If you are ok with it, consider to purchase a few tantalum caps to replace out the ceramics on your board OR search out a drop in replacement for these LDOs that are ceramic cap friendly. We use parts from DIODES.COM who has purchased many of the offshore vendors and they offer such LDOs. Let me check their website now.

[mon2]

Ok, see here:

SPX1117 Data Sheet (mouser.com)

* this datasheet notes the part is ok with ceramic or tantalums. You may be ok with the LDO you already have with the ceramic but in general, the datasheet must note that it is compatible with the low ESR ceramic parts

So, Mouser shows this in stock:

(we use Maxlinear on some of our designs with ceramic caps - no issues)

SPX1117M3-L-1-5/TR MaxLinear | Mouser
; 1v5 output fixed
; compatible with the pinout of your adapter board
; compatible with ceramic caps


Microsoft Word - AP2114 D2.2 130114 (mouser.com)
; 1v2 output fixed
; compatible with the pinout of your adapter board
; compatible with ceramic caps

AP2111H-1.2TRG1 Diodes Incorporated | Mouser

[baglaivlad]

Thank you!

This is how 1.8V output looks like, please see the attachment.

I can borrow some tantalum caps from MacBook boards, if needed.

[mon2]

No issue here. Your LDO is most likely ok with the use of the caps onboard.

[baglaivlad]

Good to know, thank you! Will test purchased LDOs as well.